INTRODUCTION

Alpha-fetoprotein (AFP) is a cancer-associated fetal glycoprotein, normally produced in the fetal liver and yolk sac, and its high serum level is a useful clinical marker for hepatocellular carcinoma and yolk sac tumor. Although the physicochemical and structural properties of this 62-72 kDa glycoprotein have been largely documented, only in vitro functional roles of this oncofetal protein have been ascertained to date. Such physiological properties of the oncofetal protein have encompassed mainly ligand carrier/transport function and modulation of in vitro immune response assays. In the last decade, the growth regulatory properties of AFP have aroused interest as a result of studies involving ontogenetic and oncogenic growth in both cell culture and animal models1, 2, 3. A myriad of studies have now described that AFP is capable of regulating growth in ovarian, placental, urerine, hepatic phagocytic, bone marrow, and lymphatic cells4 in addition to various neoplastic cells5. AFP should no longer be considered merely a fetal form of albumin only to be employed as a marker for cancer and fetal disorders. Rather, AFP should now be considered as a possible direct or indirect factor associated with the regulation of growth, differentiation, regeneration, and transformation in both ontogenetic and oncogenic growth processes. However, little information regarding the biochemical features of alpha-fetoprotein determined using molecular biological techniques is available. Although it is currently thought that a 62- to 67-kDa membrane protein on the surface of monocytes and phagocytes is specific for AFP6, 7, the properties of the binding sites were still unknown in most tumor cell lines. Furthermore, few studies have focused on its intracellular signaling events and gene expression. The goal of this study was to characterize the AFP receptor, its possible signal transduction pathway and its proliferous functions in human NIH 3T3 cells.

MATERIALS AND METHODS

Reagents

The cAMP CPBA kit and Na[125I] were purchased from Amersham, UK. Fluo-3 AM ester was from BioRad (USA). Purified AFP was from Sigma (USA). Monoclonal antibody for K-ras P21 protein was purchased from Maxim Biotech, Inc (USA).

Isolation and identification of human AFP

Human AFP was prepared by the method as described elsewhere8. Briefly, human cord blood AFP was precipitated by ammonium sulphate and passed through an anti-AFP affinity chromatography column. AFP-positive fractions were collected and concentrated. The purity of prepared AFP was 92.7% determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The protein was stored at −80 °C until use.

Preparation of monoclonal antibody against AFP

Monoclonal antibody against AFP (Anti-AFP) was prepared according to the routine procedure9. In short, BABL/C mice was immunized with purified AFP (Sigma) in complete Freunds adjuvant at 2 to 3 w intervals. The spleen cells separated from mice were fused with myeloma cells to form a stable antibody-producing hybridoma cell line. Positive clones were screened by the method of ELISA and inoculated into BALB/C mice. The anti-AFP was harvested from ascites fluid and purified with affinity chromatography. The specificity of monoclonal antibody with a titers higher than 5000 for AFP was ascertained by ELISA and Western blotting assay to obviate the interference of albumin which had a similar structure to AFP. The results in ELISA and Western blotting assay showed the specific binding of monoclonal antibody to AFP and negative reaction to albumin(data not show).

Measurement of the cellular proliferation

NIH 3T3 cells were suspended in RPMI-1640 medium (containing 10% fetal calf serum) and added into 24-well plate at 1 ml per well followed by incubating at 37°C in a humidified atmosphere of 5% CO2 for 48 h. The supernatant was removed and replaced with 150 μl of fresh medium (without fetal calf serum) for another 24 h. Different concentrations of AFP (0-80 mg/L), human serum albumin (HSA) or Anti-AFP were administrated into each well respectively for 20 h and then pulsed with 1 mCi of 3H-TdR. The cells were harvested 4 h later onto glass microfiber filter using a multiple sample harvester. The incorporation of 3H-TdR was measured by using LKB 1209 Rackbeta liquid scintillation counter. To determine whether the influence of AFP on the proliferation was specific, the blockage of Anti-AFP and HSA as a structure analog were also used as controls.

AFP receptor binding assay

NIH 3T3 cells were maintained in suspension cultures at 37°C in RPMI-1640 medium without serum for 12 h and washed three times with cold medium. Resuspended cells were passed through a 300 mesh screen and adjusted to 1×106 cells per ml. 125I-AFP was prepared by the iodogen method and run through a column of Sephadex-G25 to remove some of the unincorporated 125I. The specific activity of 125I-AFP was 2715 Ci per mmol and the radiochemical purity of 125I was 99.4%. Each reaction contained 7× 105 cells, 125I-AFP of 10−15 M (5×104 cpm) and different concentrations of non-labeled AFP (0.25-64.5 ng). The reactions were triplicated and performed at 4 ° for 2 h. Nonspecific binding was determined using 2 μg.ml-1 partially purified AFP (90% purification) per well. All samples were collected onto glassfiber (saturated with 0.5 % albumin in advance) and washed three times with 15 ml of PBS. The radioactivity of 125I was detected by a -counter.

Extraction and detection of cAMP

Cells were adjusted to 4×104 cells per ml with medium (containing 10% fetal calf serum) and aliquot into 24 well plates to incubate for 24 h. The supernatant was removed and resuspended in the medium supplemented with 0.1% egg albumin and 2.5×10−2 M of HEPES and 2 mM IBMX (3-methyl-1-isobutyl-xanthine) at 37°C for 15 min. AFP (20 mg/L) or Anti-AFP (40 mg/L) were administrated into each well respectively for 4 h. Extraction of cAMP was performed according to the method described by Iwashita10. In short, the supernatant was removed and replaced with 1 ml of cold PBS per well. After wash 3 times, the pellet was frozen in −80°C for 30 min and then added 0.5 ml of HCl (0.05 N) into each well for another 30 min in −80°C.The samples were melted in room temperature and transferred into eppendorf tubes. The supernatant for each sample was taken out following the spin at 10000 g for 5 min and lyophilized. The content of cAMP was determined by the radio-immunoassay according to the instruction of cAMP detection kit.

Determination of protein kinase A activity

NIH 3T3 cells were adjusted to a final concentrations of 4×104 cells per ml with RPMI-1640 (containing 10% fetal calf serum) in culture plates for 48 h and then resuspended in medium without serum for 24 h. The cells were administrated with AFP (20 mg/L), Anti-AFP (40 mg/L) or AFP (20 mg/L) plus Anti-AFP (40 mg/L) respectively. After being treated for 2, 6, 12 and 24 h, the cells were washed and resuspended in 1 ml of PBS (0.15 M, pH 7.4). The measurement of PKA activity has been carried out as described in Park et al with modifications11. Briefly, 40 μl of extracted enzyme fractions were mixed with 160 μl of the solution at the final concentration of 2.0×10−2 M Tris-HCl (pH 7.5), 5×10−3 M MgCl2, 0.25 g/L BSA, 0.5g/L histone, 2×10−7 M ATP(32P ATP, 3.7×104 Bq) and 8.0×10−6 M cAMP at 37°Cfor 10 min. After being followed by incubation on ice for 5 min to terminate the reaction, 150 μl of the solution from each sample was collected onto Whatmen GF/C filter paper. After washing 2× with 10% TCA-2% phosphoric acid for 30 min at room temperature followed by 2× wash with 5% TCA for 30 min, the activities of PKA were measured by liquid scintillation counter and expressed as pmol value of 32P in histone catalyzed by per mg protein per min.

Western blot immunodetection

1×105 cells per ml were maintained in RPMI-1640 medium without serum at 37°Cfor 12 h. Then, the cells were treated with AFP (20 mg/L), Anti-AFP (40 mg/L) or AFP (20 mg/L) plus Anti-AFP (40 mg/L) respectively for 24 h. After three times wash, the cells in each reaction were incubated in 10 μl of lysis buffer containing 0.2% Triton X-100, 0.5 M NaCl, 0.5 M sucrose, 1×10−3 M EDTA, 1.5×10−4 M specimine, 5×10−4 M spermidine, 1.0×10−2 M HEPES (pH 8.0), 2.0×10−4 M phenylmethylsulfonyl fluoride, 2 mg leupeptin/ml, 2 μg pepstatin/ml, 24 IU aprotinin/ml and 7×10−3 M β-mercaptoethanol. 40 mg of total proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membrane for immunodetection. SDS-PAGE molecular weight markers (Bio-Rad) verified the correct location of the visualized bands. The membranes were blocked in 5% carnation nonfat milk (w/v) in PBS-Tween, probed with anti-K-ras P21 antibody and followed by second antibody (Goat anti-mouse Ig-alkaline phosphatase). Immunoreaction protein was detected by the color develop system (NBT/BCIP). Quantitative analysis of bands were performed by an automated digitizing system (UN-SCAN-IT, Silk Scientific, Inc.) and expressed with pix total.

Statistical analysis Data were expressed as mean±s obtained from at least 3 or 4 independent experiments and analyzed by t test.

RESULTS

The effect of AFP on the cellular proliferation

After treatment of NIH 3T3 cells with AFP for 24 h, the proliferation of NIH 3T3 cells could be markedly enhanced in a dose dependent manner. The extents of increase were up to 21.0-121.9 % compared with the control group (Fig 1). HSA could not show the obvious influences on the proliferation. Although Anti-AFP alone did not change the incorporation of 3H-TdR, the effect of AFP could be reversed by Anti-AFP (Fig 2).

Figure 1
figure 1

The effects of different concentrations (0-80 mg/L) of AFP or HSA on the incorporation of 3H-TdR into DNA in NIH 3T3 cells. n=6, *P<0.05 and **P<0.01 vsdosage 0 mg/L.

Full size image
Figure 2
figure 2

The blockage effects of Anti-AFP to AFP on the proliferation of NIH 3T3 cells. Treated groups: 1. control; 2. AFP (20 mg/L); 3. Anti-AFP (40 mg/L); 4. AFP (20 mg/L) + Anti-AFP (40 mg/L); 5. HSA (20 mg/L). n=6; ** P<0.01 vs control group.

Full size image

Distribution of AFP receptor on the membranes of NIH 3T3

The binding sites of AFP on the surface of the cell and KD values were calculated based on the Scatchard plot analysis of 125I-AFP. According to the Scatchard analysis, there were two classes of receptors with KD 2.722×10−9 M (with 12810 sites per cell) and KD 8.931×10−8M (119700 sites per cell) (Fig 3).

Figure 3
figure 3

Scatchard analysis of 125I-AFP binding to NIH 3T3 cells. The properties of AFP receptor was detected with receptor binding assay and analyzed by a program of Radioligand Binding Assay of Receptor. The data was selected from three independent experiments.

Full size image

Determination of intracellular cAMP concentration

AFP could markedly elevated the concentrations of cAMP up to 267% in NIH 3T3 cells (Fig 4). Anti-AFP could not alter the concentrations of cAMP, but it could reverse the effect of AFP. HSA did not influence the concentration of cAMP.

Figure 4
figure 4

The effects of AFP (20 mg/L), Anti-AFP (40 mg/L), AFP (20 mg/L) + Anti-AFP (40 mg/L) or HSA (20 mg/L) on the cAMP concentration of NIH 3T3 cells. Data presented as mean values of 8 samples±s. **P<0.01 vscontrol.

Full size image

Detection of PKA activity

The activities of PKA in the cytosol of NIH 3T3 cells was obviously elevated after treated with AFP for 2, 6, 12 or 24 h (Fig 5). The activities of PKA were respectively increased up to 14.8, 81.1, 63.3 and 14.9% at each time point. The peak value was achieved at 6 h and then declined gradually, but still maintained a higher activity for several hours. Anti-AFP monoclonal antibody and HSA did not obviously influence the activity of PKA, but Anti-AFP could block the effects of AFP on the activity of PKA.

Figure 5
figure 5

The effects of AFP (20 mg/L), Anti-AFP (40 mg/L), AFP (20 mg/L) + Anti-AFP (40 mg/L) or HSA (20 mg/L) on the PKA activity of NIH 3T3 cells. *P<0.05 and **P<0.01 vs control. Data presented as mean values of 4 samples±s.

Full size image

Determination on the expression of K-ras P21 proteins

The result in Fig 6 demonstrated the overexpression of K-ras P21 protein in AFP-treated group in NIH 3T3 cells at different treated time (Fig 6). The upregulated effects of AFP on the expression of K-ras p21 gene could be blocked by Anti-AFP (Fig 7). HSA could not influence the contents of this protein. Each figure was selected from 3 similar results.

Figure 6
figure 6

The cells were treated with HSA (20 mg/L) or AFP (20 mg/L) for 6, 12 or 24 h. The effect of AFP on the expression of K-ras P21 proteins in NIH 3T3 cells was detected by Western blot assay. Relative intensity for each band was determined by an automated digitizing system. The data was selected from four similar experiments.

Full size image
Figure 7
figure 7

The effects of AFP on the expression of K-ras P21 proteins in NIH 3T3 cells were detected by Western blot assay. Relative intensity for each band was determined by an automated digitizing system. Treated groups: Control, control group; HAS, HSA treated group; AFP, AFP treated group; Anti-AFP, Anti-AFP treated group; AFP +Anti-AFP, AFP plus Anti-AFP treated group. The data was selected from four similar experiments.

Full size image

DISCUSSION

AFP, as an oncodevelopmental gene product which is expressed at high levels in the adult during liver regeneration and hepatocarcinogenesis, has been recently revealed as a protein with growth factor-like. Although the biological regulation of AFP on the cell growth has been extensively evaluated8, 12, 13, 14 and verified by our results of the cellular proliferation assay, neither the properties of AFP receptor in tumor cells were fully clarified, nor the subsequent events was tested after AFP binding the receptor. Some recent investigations of AFP-receptor have emerged which documented the existence and further characterized the nature of such binding protein. Our data revealed that there were two classes of receptors in cultured NIH 3T3 cells. Scatchard analysis displayed the presence of two specific AFP binding site classes with KD of 10−9 and 10−8 M, which was consistent with some similar experiments that characterized the KD of binding protein in the monocytes were at a range of 10−11–10−7M6, 15.

Little information about the effect of AFP on the signal transduction was available to explain the intracellular mechanism. In the present experiment, the evidences that the activities of PKA and the content of cAMP in the NIH 3T3 cells were signi- ficantly elevated indicated a cAMP-dependent protein kinase pathway was involved in the effects of AFP on the tumor cells. Although the albumin is similar to alpha-fetoprotein in structure and believed to be derived from a common ancestral gene, HSA was not capable of altering the activities of PKA and the content of cAMP. In all our AFP studies, none of the results showed that HSA as a control was able to alter the parameters of cell proliferation although it can non-specifically bind to the cell surface. Moreover, monoclonal antibody against AFP could significantly reverse the effect of AFP, which implied that the accruing events in the intracellular signal transduction following binding of AFP to its receptor were specific.

Although growing evidences have verified the effects of AFP on the growth of tumor cells, little works were focused on the subsequent events in nucleus. The impacts of overexpression of protooncogenes on tumor growth have been largely documented. The previous works in our laboratory have been also shown an overexpression of p53 gene (data not shown). The other laboratory also demonstrated the alteration of related genes in malignant transformation of NIH 3T3 cells16. All there evidences indicated the alteration of expression of related genes. In this investigation, the expression of K-ras p21 gene was determined for its responsibility in AFP treatment. It was prevalent accepted that K-ras P21 activates raf, which in turn passes the signal through the intracellular signal regulated kinases to stimulate cell division, and that this pathway is upregulated when K-ras is mutated17. In fact, the alterations of K-ras p21 gene expression have been largely documented and related to tumor growth18, 19, 20. In the present experiment, K-ras P21 protein was over produced in the presence of AFP. The finding indicates that the cell behavior induced by AFP was closely coupled with the change of K-ras gene expression. Thus, it thereby implied the mechanism by which the AFP-induced oncogenesis might be mediated through K-ras P21 signal pathways.

Collectively, the functionary mechanism of AFP on the tumor growth may be attributed, at least in part, to the receptor-mediated intracellular communication and the change of relative genes expression. Some other unknown factors which may also play key role in the effects of AFP need to be further characterized. Further investigations on the signal pathway will shed further light on the mechanism of AFP action.